Reverse Regression Research Articles

Effect of Sildenafil and Rolipram on Adrenergic Responses in Isolated Human and Monkey Corpus Cavernosum

ObjectiveEvaluate and compare effects of phosphodiesterase inhibitors (PDEIs), sildenafil and rolipram, on adrenergic contractile responses of human and monkey cavernosal smooth muscle. MethodsHuman penises were obtained from patients undergoing gender reassignment surgery. Isolated human and monkey corpus cavernosum (CC) strips were suspended in tissue bath chambers for isometric tension experiments. The effects of the drugs on precontracted monkey and human CC and neurogenic contractions in human CC were investigated. ResultsBoth sildenafil and rolipram induced concentration-dependent relaxations of human and monkey CC strips precontracted with noradrenaline (NA). The IC50 values, determined by reverse regression for nitroglycerin (NTG), isoprenaline, and sildenafil in monkey CC, were, respectively, 1.5±0.9×10−7M, 3.7±0.6×10−6M, and 1.7±0.7×10−5M. Similarly, in human CC muscle, sildenafil was weaker than NTG as a muscle relaxant. Sildenafil, 1.5μM, reduced neurogenic contractions in human CC due to stimulation of predominantly adrenergic nerves. The suppressant effect of sildenafil on adrenergic transmission was attenuated in CC strips pretreated with Nω-nitro-l-arginine and overcome with a higher stimulus frequency or tetraethylammonium. Rolipram partially inhibited adrenergic excitatory response but without significantly affecting NA-induced contraction. ConclusionsSildenafil and rolipram induced concentration-dependent reversal of human and monkey CC tone mediated by NA. Both PDEIs attenuated contractile adrenergic response of human CC to electrical stimulation. The results also underline the importance of the cyclic adenosine monophosphate-dependent signalling pathway in regulating the tone. PDE4 inhibition in CC is an additional mechanism for erection and potential therapeutic adjunct.

Putting out Fires: An Examination of the Determinants of State Clean Indoor-Air Laws

1. IntroductionFew products have generated as much public scrutiny as tobacco. In light of the health consequences associated with smoking, the federal government took several steps in the 1960s (such as published health warnings and restrictions on cigarette advertising) to reduce the incidence of smoking. Beginning in the 1970s, state governments began to play an increasing role in the anti-smoking campaign by adopting clean indoor-air laws, which restricted smoking in particular locations, such as restaurants, bars, and work sites.1Various studies, including Wasserman et al. (1991), Chaloupka (1992), Chaloupka and Saffer (1992), and Yurekli and Zhang (2000), generally confirmed that clean indoor-air laws are effective at reducing cigarette demand. However, the nature of the smoking restriction appears to matter. In particular, Chaloupka (1992) and Chaloupka and Saffer (1992) found that restrictions on smoking in public places reduced cigarette demand, whereas restrictions at private work sites do not significantly reduce demand.2Due to variations in antismoking laws across states, only a few studies have addressed the demand for clean indoor-air laws. In a survey of individuals from San Luis Obispo, California, for example, Boyes and Marlow (1996) found nonsmokers and women to be more supportive of a smoking ban in restaurants and bars. Alternatively, Chaloupka and Saffer (1992) used state-level data and found that differences in smoking restrictions in public places and at private work sites depend on several factors, including cigarette prices, income, tobacco production, religious affiliation, and political activity. More recently, Hersch, Del Rossi, and Viscusi (2004) found that voting preferences of state residents and the political affiliation of lawmakers were the main factors that affect whether or not a restriction is adopted at a specific location (for example, restaurants, bars, malls, enclosed arenas, and hospitals). However, smoking habits could have an effect on voting preferences; therefore, these results could show the indirect effect of tobacco consumption on smoking restrictions.There are several limitations of the existing literature that lead to a need for further analysis. First, studies have ignored the potential role of cigarette taxation as a determinant of smoking bans. For example, if states adopt a general antismoking position, taxes could be used in conjunction with smoking bans to reduce tobacco consumption. Thus, depending on the argument, states may view taxation as a policy substitute or a policy complement to clean indoor-air laws. Second, the potential endogeneity of right-hand-side variables has not been addressed. For example, tobacco consumption is often included as a determinant of smoking bans; however, tobacco consumption may potentially be endogenous. In general, studies have found that restrictions on smoking in public places reduce cigarette demand, which implies that cigarette consumption is potentially endogenous in the reverse regression of smoking on restrictions. It is unfortunate that the literature has not addressed this possibility in a systematic manner.3 Third, except for Hersch, Del Rossi, and Viscusi (2004), studies have failed to account for differences in the demand for clean indoor-air laws across different restricted locations.This study addresses these limitations by estimating the demand for clean indoor-air laws using a panel of state-level data. By using a panel data set, we were able to control for state-specific effects that made it more or less likely that a given state had a smoking ban in a particular area. Specifically, we accounted for the potential endogeneity of cigarette consumption and taxation within a random-effects Probit model of the adoption of smoking restrictions across six locations--public places, government buildings, private work sites, schools, health care facilities, and restaurants. …

Linkage mapping of a complex trait in the New York population of the GAW14 simulated dataset: a multivariate phenotype approach

Multivariate phenotypes underlie complex traits. Thus, instead of using the end-point trait, it may be statistically more powerful to use a multivariate phenotype correlated to the end-point trait for detecting linkage. In this study, we develop a reverse regression method to analyze linkage of Kofendrerd Personality Disorder affection status in the New York population of the Genetic Analysis Workshop 14 (GAW14) simulated dataset. When we used the multivariate phenotype, we obtained significant evidence of linkage near four of the six putative loci in at least 25% of the replicates. On the other hand, the linkage analysis based on Kofendrerd Personality Disorder status as a phenotype produced significant findings only near two of the loci and in a smaller proportion of replicates.

Pregnancy loss in dairy cattle in the Waikato region of New Zealand

AIM: To define the incidence rate of pregnancy loss and risk factors for those losses in pasture-fed dairy cattle in the Waikato region of New Zealand. METHODS: Cows (n=2,004) from 10 pasture-fed, spring-calving dairy herds in the Waikato were enrolled following confirmation of pregnancy 29–45 days after insemination, for inseminations that occurred within the first 16 days of the seasonal breeding period. Transrectal ultrasonographic examinations for pregnancy were conducted at approximately 6, 8, 10, 14 and 22 weeks gestation, and subsequent calving data were recovered. Pregnancy loss was defined as having occurred when a confirmed pregnancy was not rediagnosed, when gross abortion was detected, or when a cow calved <265 days after the confirmed conception date. Data were analysed using reverse stepwise logistic regression and Cox's proportional hazards analysis. RESULTS: A total of 128 (6.4%) pregnancy losses were detected. The incidence rate was higher in early compared to late gestation (10.9 vs 2.8 losses/10,000 cow-days between Weeks 6–10 vs Weeks 10–14, respectively; p<0.001). Higher rates of loss were associated with the occurrence of clinical mastitis (Hazards ratio (HR)=1.57; p=0.071), being treated for anoestrus (HR=1.69; p=0.007), and in cows that had calving-to-conception intervals ≤63 days compared with those that had calving-to-conception intervals >92 days (HR=2.49; p=0.06). In addition, the rate of pregnancy loss differed between herds (p=0.05). CONCLUSIONS: The highest rate of pregnancy loss occurred in early gestation. Clinical mastitis, anoestrus and calving late in the calving season were risk factors for pregnancy loss. CLINICAL RELEVANCE: Pregnancy diagnosis using ultrasonography can be undertaken from 28 days post-insemination. However, due to the high rate of pregnancy loss at this stage of gestation, herdowners need to be warned of possible losses, and cows should be re-examined to confirm pregnancy before certification of pregnancy status is given.

Computational Models for Prediction of IVF/ICSI Outcomes With Surgically Retrieved Sperm

Various factors may be involved in determining the outcome of IVF/ICSI. Using several linear and nonlinear mathematical models, we investigated the correlation of maternal age, type of male factor infertility, sperm retrieval techniques and sperm type with IVF/ICSI outcome. A dataset of 113 exemplars containing input features maternal age, sperm retrieval technique (testicular sperm extraction, testicular sperm aspiration and microscopic epididymal sperm aspiration), type of sperm used (cryopreserved or “fresh”), type of male factor (previous vasectomy with or without subsequent reversal, congenital absence of the vas deferens, other obstructive conditions, varicocele, non-obstructive azoospermia, miscellaneous) and output intra-uterine pregnancy achieved by IVF/ICSI was randomized into a modeling (“training”) set of 83 and cross-validation (“test”) set of 30, with similar outcome frequencies preserved in each set. We used neUROn++, a set of C++ programs we developed using the Cygwin (Red Hat) GNU C++ port for Windows (Microsoft) distributed across Pentium (Intel) platforms, to model the dataset using the linear methods linear and quadratic discriminant function analysis (LDFA, QDFA) and logistic regression (LR), and the nonlinear method of neural computation (NNET) with several investigated architectures. A 4 hidden node (HN) network was found to have the highest accuracy, compared to LDFA, QDFA and logistic regression, with a ROC area for the test set of 0.783. Reverse regression of the neural network based on Wilk’s Generalized Likelihood Ratio test revealed maternal age to be most significant (p=0.025) feature in predicting pregnancy, followed by sperm type (p=0.076). Type of male factor (p=0.47) and sperm retrieval technique (p=0.88) did not predict outcome. Tabled 1 We have investigated linear and nonlinear computational models of IVF/ICSI. The demonstration that IVF/ICSI-induced intra-uterine pregnancy correlates to maternal age and to the type of sperm used supports the use of our neurocomputational model as a predictive tool when planning for the use of assisted reproductive techniques in the presence of male factor infertility.

Predicting hypogonadism in men based upon age, presence of erectile dysfunction, and depression

Hypogonadism, a disorder associated with aging, can cause significant morbidity. As clinical manifestations of hypogonadism can be subtle, the challenge and the burden of diagnosis remain the responsibility of the clinician. Four different analytic methods were used to predict hypogonadism in men based upon age, the presence of erectile dysfunction (ED) and depression. 218 men were classified by age, serum testosterone level, the presence of ED and depression. Depression was determined by the Center for Epidemiologic Studies Depression Scale (CES-D). ED was assessed by the Sexual Health Inventory for Men (SHIM). Hypogonadism was defined as a serum testosterone level <300 ng/dl. An artificial neural network (ANN) was programmed and trained to predict hypogonadism based upon age, SHIM, and CES-D scores. Subject data was randomly partitioned into a training set of 148 (67.9%) and a test set of 70 (32.1%). The ANN processed the test set only after the training was complete. The discrete predicted binary output was set to (0) if testosterone level was <300 ng/dl or (1) if >300 ng/dl. The data was also analyzed by standard logistic regression (LR), linear and quadratic discriminant function analysis (LDFA and QDFA, respectively). Reverse regression (RR) analysis evaluated the statistical significance of each risk factor. The ANN can accurately predict hypogonadism in men based upon age, the presence of ED, and depression (receiver-operating characteristic=0.725). A four hidden node network was found to have the highest accuracy. RR revealed the depression index score to be most significant variable (P=0.0019), followed by SHIM score (P=0.00602), and then by age (P=0.015). Hypogonadism can be predicated by an ANN using the input factors of age, ED, and depression. This model can help clinicians assess the need for endocrinologic evaluation in men.

COMPUTATIONAL MODELS FOR DETECTION OF ERECTILE DYSFUNCTION

Vascular comorbidities are well known to correlate with erectile dysfunction (ED) but a correlation with hypogonadism and depression is less clear. Using several linear and nonlinear mathematical models we investigated the correlation of age, hypogonadism and depression with ED using the Sexual Health Inventory for Men (SHIM) as a surrogate marker for ED. A data set of 140 exemplars containing the input features age, total testosterone in ng/dl, Center for Epidemiologic Studies Depression scale score and output SHIM score (thresholded to 10 for moderate ED) was randomized into a modeling (training) set of 105 and a cross-validation (test) set of 35 with similar outcome frequencies preserved in each set. Using neUROn++, a set of C++ programs that we developed using the Cygwin (Red Hat, Raleigh, North Carolina) GNU C++ port for Windows (Microsoft, Redmond, Washington) distributed across Pentium (Intel, Santa Clara, California) platforms we modeled the data set using the linear methods, linear and quadratic discriminant function analysis, and logistic regression, and the nonlinear method of neural computation with several investigated architectures. A 4 hidden node network was found to have the highest accuracy compared to linear and quadratic discriminant function analyses, and logistic regression. ROC areas for the test set were 0.702, 0.645, 0.676 and 0.618, respectively. Analysis of the neural network demonstrated that moderate ED correlated with patient age and depression score. Forward and reverse regression of the neural network based on Wilk's generalized likelihood ratio test revealed that age was most significant (p <0.001), followed by Center for Epidemiologic Studies Depression Scale score (p <0.03), followed by testosterone (p >0.6). We investigated linear and nonlinear computational models of moderate ED. To our knowledge this is the first demonstration that SHIM correlates with age and a depression metric. Furthermore, moderate ED based on SHIM with a correlation with age may now provide a rationale and basis for future investigation into the understanding of age related erectile pathophysiology.

Computational models for prediction of IVF/ICSI outcomes with surgically retrieved spermatozoa

IVF/intracytoplasmic sperm injection (ICSI) using surgically retrieved spermatozoa (SRS) is a key option in the treatment of severe male infertility. It was aimed to develop a computational model for the prediction of this modality's outcome. A dataset of 113 exemplars, derived from patients who underwent IVF/ICSI with SRS, was retrospectively analysed. The dataset, containing input features maternal age, sperm retrieval technique, type of spermatozoa used, type of male factor and output intrauterine pregnancy, was randomized into a modelling (‘training’) set of 83 and cross-validation (‘test’) set of 30. neUROn++, a set of C++ programs, was used to model the dataset using linear and quadratic discriminant function analysis, logistic regression, and neural computation. A 4-hidden node neural network was found to have the highest accuracy, with a test set receiver operator characteristic (ROC) curve area of 0.783. Reverse regression of this neural network showed maternal age to be the most significant feature in predicting pregnancy ( P = 0.025), followed by sperm type ( P = 0.076). Type of male factor ( P = 0.47) and sperm retrieval technique ( P = 0.88) did not predict outcome. In summary, a neural network of clinical relevance was found to be superior in terms of IVF/ICSI outcome prediction. Future media deployment is planned.

A neurocomputational model for prostate carcinoma detection.

Current guidelines for prostate carcinoma screening rely primarily on the digital rectal examination (DRE) and prostate specific antigen (PSA). Well described patient risk factors for prostate carcinoma also include age, ethnicity, family history, and complexed PSA. However, due to the nonlinear relation of each of these variables with prostate carcinoma, it is difficult to predict reliably each patient's risk based on linear univariate analysis. The authors investigated a neural network to model the risk of prostate carcinoma by seven readily available clinical features. The database for the current study comprised 3268 men recently evaluated for the early detection of prostate carcinoma. The seven clinical features evaluated included age, race, family history, International Prostate Symptom Score (IPSS), DRE, and total and complexed PSA. Three hundred forty-eight subjects in the dataset included men with determined prostate biopsy outcomes and for whom at least 6 of 7 features were available. The dataset was divided randomly into a training set (60%) and a test set (40%), with n1/n2 cross-validation used to evaluate model accuracy, and was modeled with linear and quadratic discriminant function analysis and a neural computational system. After a model with acceptable goodness of fit was achieved, reverse regression analysis using Wilks's generalized likelihood ratio test was performed to evaluate the statistical significance of each input variable. The receiving operating characteristic (ROC) area for the neural computational system in the test set was 0.825, whereas total PSA and complexed PSA alone had ROC areas of 0.678 and 0.697, respectively. The ROC area of logistic regression in the test set was 0.510, linear discriminant function analysis was 0.674, and quadratic discriminant function analysis was 0.011. All were significantly less than the ROC area of the neural computational model (all Ps < 0.002). Reverse regression based on Wilks's generalized likelihood ratio test demonstrated each input feature to be highly significant to the model (all Ps << 0.000001). The authors modeled a combination of well described patient risk factors for prostate carcinoma using a neural computational system with acceptable goodness of fit. They demonstrated that each of the seven variates on which the model was based was critically significant to model performance. The authors presented this model for clinical use and suggested that clinicians use it in deciding to perform prostate biopsy.

Reverse Regression and Orthogonal Regression in Employment Discrimination Analysis

Abstract No abstract available.

A neural computational model of ICSI outcomes from sperm source and other features

Objective: With the success and increasing popularity of ICSI, accurately predicting outcomes from this artificial reproductive technique becomes highly desirable. We examined a large dataset of ICSI outcomes and associated clinical features such as sperm source, which technician performed the procedure, number of ova retrieved and maternal age, and modeled outcomes using neural computational techniques. Neural computation employs computer programs that implement mathematical models based loosely on the function of networks of biological neurons. Statistical feature extraction, allowing examination of the significance of individual features such as sperm source, was employed using Wilk’s Generalized Likelihood Ratio Test (GLRT). Design: Neural computational modeling of retrospective data. Materials/Methods: We modeled fertilization outcomes from ICSI from technician (7 different technicians performed the procedure,) sperm source (testis extraction, MESA, EEJ,) number of ova retrieved and maternal age using neUROn2++, a set of C++ programs we developed to implement neurocomputational algorithms, discriminant function and regression techniques using the Cygwin (Red Hat) GNU C++ port for Windows (Microsoft) and Visual C++ (Microsoft) distributed across Pentium (Intel) platforms. Wilk’s GLRT was implemented in neUROn2++ for statistical feature extraction. A dataset of 1070 unique ICSI outcomes was randomized into a training set of 1349 exemplars and 449 test exemplars. N1/N2 cross-validation was employed. Results: Overall model accuracy was ROC AUC 0.922 (sensitivity 86.9%, specificity 79.0%) in the training set, and 0.897 (sensitivity 82.8%, specificity 75.1%) in the test set. Feature extraction using Wilk’s GLRT in reverse regression is shown in the table. All features were significant to the model (all p <1e-5), however, with p <1e-106, technician was highly significant.TableRegression analysis of neural computational model to predict ICSI outcomes.Featurep-valueTechnician1.04e-107No. ova retrieved9.84e-45Testis extraction3.81e-21Maternal age7.70e-20MESA2.33e-10EEJ1.52e-06p-values indicate significance of the individual feature to the likelihood of fertilization with ICSI. Open table in a new tab p-values indicate significance of the individual feature to the likelihood of fertilization with ICSI. Conclusions: We modeled ICSI outcome with high accuracy (ROC AUC 0.897 in the test set) using a neural computational approach. In order to determine the significance of individual features to the model’s outcome, regression based on Wilk’s GLRT was performed. All features examined (sperm source, number of ova retrieved, maternal age and which technician performed the procedure) were statistically significant to the model’s outcome. Supported by: NIH P01 HD36289 to Dolores J Lamb, Larry I Lipshultz and Craig Niederberger.

Least squares estimators in measurement error models under the balanced loss function

The ultrastructural form of the measurement error model is considered and a comparison of the direct and the reverse regression estimators is made under the balanced loss function, which explicitly takes into account the accuracy of the predictions and the precision of estimation.

Applying reverse regression techniques in earnings–return analyses

Measurement error in unexpected earnings is recognized as a source of bias in examinations of the relation between earnings and returns. Reverse regression procedures are commonly used as a means of coping with this bias. This study examines the properties of reverse regression procedures in multi-interacted variable settings with a specific focus on the earnings response coefficient (ERC) analysis of Collins and Kothari (J. Account. Econom. 11 (1989) 143.). It shows that both conventional reverse regression techniques and novel techniques employed by Collins and Kothari are not robust. It also demonstrates how reverse regression techniques can be successfully employed in such settings using non-interacted-variable designs.

Forward versus reverse regression and cointegration

In this paper we investigate the conditions under which the ‘forward’ dynamic regression of y on x is equivalent to the ‘reverse’ dynamic regression of x on y in the sense that the long-run coefficient obtained from the forward regression is equal to the reciprocal of the long-run coefficient from the reverse regression. We show that this equivalence holds if and only if the two series are cointegrated.

The Use Of “Reverse Regression” In Employment Discrimination Analysis

Abstract No abstract available.

Asymptotic efficiency properties of least squares in an ultrastructural model

This paper considers the estimation of both the intercept term and the slope parameter in a linear ultrastructural model using direct regression and reverse regression methods. Without assuming the errors to be normally distributed, asymptotic expressions for the bias and the variance of the estimators are derived and analyzed. The effect of departure of the errors from normality is also studied.

Application of Reverse Regression to Boston Federal Reserve Data Refutes Claims of Discrimination

The topic of mortgage discrimination has received renewed interest since publication of the Boston Federal Reserve Bank study based on 1990 Home Mortgage Disclosure Act data. That study used traditional direct logistic regression to assess the influence of race on the probability of mortgage loan denial and reported the parameter estimate of race to be positive and significantly different from zero across several model specifications, thereby supporting contentions of discriminatory behavior. This paper develops an alternate approach, reverse regression, a method often used in the measurement of gender discrimination in labor markets. After discussion of theoretical issues regarding model choice, results of a reverse regression on the Boston Federal Reverse Bank study dataset are reported. Contrary to results using direct methods, reverse regression does not support contentions of mortgage discrimination in the Boston mortgage market. Rather, the lower overall qualifications of minority applicants are lik...

The Convergence of Labour Productivity in British and German Industries

When a Galtonian cross-section regression model is used to analyse data on the relative labour productivity of 30 British and German manufacturing industries 1960-1989, it is found that there are strong tendencies for labour productivity in the two countries to converge. This is particularly true for 1960-73 and for 1979-89, when there was Galtonian regression towards the (geometric) mean level of relative productivity. There was no tendency for the relative advantage or disadvantage of an industry to persist over time. When an errors-in-variables model is used, it is found that reverse regression and lagged value instrumental variable estimators do not indicate that the OLS estimators in the Galtonian model are usually biased downwards and in fact confirm the direction of the Galtonian regression.

On normality and the linear regression model

The purpose of this note is to discuss the role of normality in the context of linear[zddot]homoskedastic regression models. A new characterization result, relating the joint normal distribution and the linear, homoskedastic regression, sheds some light on the role of normality in this context. It is shown that if the assumptions of (a) linearity and (b) homoskedasticity, are supplemented with the assumption of (c) linearity of the reverse regression, assumptions (a)-(c) are tantamount to assuming joint normality of the regressors and regressands, not just conditional normality.

Determination of thallium in cement dust and sediment samples by differential-pulse anodic stripping voltammetry: a chemometric approach to linear calibration

Differential-pulse anodic stripping voltammetry (DPASV) was used to determine TI in cement kiln dust and sediment samples. The separation of TI from the sample matrix was achieved by the extraction of Tl3+ into diethyl ether from an HBr—Br2 medium. Initially, a mixture of HNO3 and HF was used to digest the cement dust samples to cause volatilization of silica as SiF4. The TI3+ was then reduced to TI+ using hydrazine sulfate prior to determination by DPASV. A calibration graph based on a deterministic relationship between the measured response y and the analyte concentration x(so that y=β0+βx+r) is also discussed using classical (or reverse) regression. The precision of parameter estimates of the model (b0 and b: ŷ=b0+bx) can be achieved by display of the diagonal elements of the hat matrix. Methods for computing the standard uncertainty in the predicted response s2ŷn and in the concentration s2x are also provided. Thallium concentrations of 700 ± 3.07, 395 ± 2.91 and 850 ± 2.59 ng g–1 were found in the cement dust and sediment samples, respectively.